Atmospheric carbonation of concrete with EOS4

Olivier Bildstein It's also entirely possible this is a physical phenomena, not just a simulation issue.  What you would want to do is make sure that when large changes in outputs start occurring, the time steps reduce automatically.  That would at least make the oscillations output more smoothly.  This is a little tricky to do in practice, since changing, say, the Courant number won't really get you there.  I would see what one of the developers like Eric Sonnenthal says about it.  By the way, could you see increase the output frequency, so we can see what those waves and oscillations look like a little better?